ONLINE FRAUD PAYMENT DETECTION USING BALANCED ML ALGORITHMS

Abstract

The rapid growth of digital payment systems and e-commerce platforms has significantly increased the risk of online fraud, making fraud detection a critical concern in modern financial systems. Fraudulent transactions not only cause financial losses but also reduce user trust in digital platforms. One of the major challenges in fraud detection is the class imbalance problem, where fraudulent transactions represent only a small fraction of the total dataset, making it difficult for traditional machine learning models to accurately detect them. This project, “Online Fraud Payment Detection Using Balanced Machine Learning Algorithms,” proposes an advanced framework that addresses data imbalance and improves fraud detection accuracy using intelligent techniques. The proposed system utilizes balanced machine learning approaches such as SMOTE (Synthetic Minority Over-sampling Technique), undersampling, and hybrid sampling methods to handle imbalanced datasets effectively. These techniques ensure that the model learns equally from both fraudulent and non-fraudulent transactions. Various machine learning algorithms, including Random Forest, Logistic Regression, Support Vector Machines (SVM), and XGBoost, are implemented and compared to identify the most effective model. Feature engineering techniques are applied to extract meaningful patterns from transaction data, such as transaction amount, time, location, and user behavior. The system also incorporates anomaly detection methods to identify unusual transaction patterns in real time. The performance of the system is evaluated using metrics such as accuracy, precision, recall, F1-score, and ROC-AUC, with a particular focus on recall and precision due to the critical nature of fraud detection. The results demonstrate that balanced learning techniques significantly improve the detection of fraudulent transactions while reducing false positives. The proposed framework provides a scalable, efficient, and reliable solution for real-time fraud detection in online payment systems. This research contributes to the development of secure financial technologies by enhancing fraud prevention mechanisms and improving user trust in digital payment platforms.